As of 16 October 2026, an updated Google privacy policy comes into force which, according to The Hacker News and Android Authority, for the first time explicitly describes the advertising implications of using the Personal Intelligence feature. The essence of the change: if a user voluntarily connects Gmail, Google Photos or Google Drive to AI tools — Search, AI Mode or Gemini — data from these services may be used for ad personalization. By default, content from Gmail, Drive and Photos is not used for advertising, but enabling cross‑service access changes this rule. Users who are not comfortable with this can opt out via the search personalization settings.
What exactly has changed in the policy
The updated policy includes wording to the effect that a user may allow certain Google services (Search, Gemini Apps) to access data from other services — Google Photos, Gmail, Drive — to provide a personalized experience. The key caveat: “Search may process some of the data you’ve granted access to in order to personalize content, including ads.” At the same time, Google emphasizes that without the user’s explicit consent, content from Drive, Gmail or Google Photos is not used for personalized advertising. These wordings are taken from the text published by The Hacker News and Android Authority; direct verification of Google’s policy page was not carried out as part of this research.
Mechanism: how Personal Intelligence creates an advertising signal
The Personal Intelligence feature, introduced by Google in January 2026, allows users to connect Gmail, Google Photos and other services to Gemini to receive personalized responses. According to Google, app connections are disabled by default: the user chooses which services to link and can disconnect them at any time. Google also states that Gemini is not trained directly on the contents of the Gmail inbox or the Google Photos library.
The practical privacy boundary lies precisely in the user’s decision to enable cross‑service access. Google’s documentation confirms that connections are disabled by default, and the advertising exception is activated only when the user themselves makes data available to services that support personalized advertising. In other words, if a user asks a question in AI Mode or Gemini with Personal Intelligence enabled, data from Gmail or Photos used to generate the response may become a signal for ad targeting.
Important nuance: disabling does not erase history
According to Android Authority, disabling a connected app does not necessarily delete information already saved in Search Services History. Data may be retained even after the original email or photo is deleted. This means that in practice protection is weaker than simply flipping a switch: data processed earlier may remain in the system.
Geography of availability
The source material states that Personal Intelligence is being rolled out in beta for users over 18 worldwide, with the exception of the European Economic Area, Switzerland, the United Kingdom and Nigeria. However, Google’s official announcements from January 2026 described the launch as a beta limited to the US, and connecting Gmail and Photos to AI Mode as an experimental Labs feature for Google AI Pro and AI Ultra subscribers in English in the US. It is likely that by October 2026 the geographic coverage had expanded, but the exact timing and conditions of expansion could not be determined from the available primary sources.
Context: OpenAI expands advertising in ChatGPT
In parallel with Google’s changes, OpenAI announced the expansion of advertising in ChatGPT to 26 new countries, bringing the total to 89. Ads are shown only to users on the free tier and the Go plan; Plus, Pro and Enterprise subscriptions remain ad‑free. OpenAI states that conversations remain closed to advertisers, user data is not sold, ads are clearly labeled, and they do not affect the chatbot’s responses. Previously, we have already written about security issues related to OpenAI products.
The control models of Google and OpenAI affect different levels of the advertising pipeline. Google decides whether personal data from connected services may inform ad targeting. OpenAI focuses on isolation: advertisers do not gain access to the content of conversations, and ads are separated from responses. Both companies declare user control, but the mechanisms are fundamentally different.
Practical recommendations
- Check connected apps: open the search personalization settings and make sure that only those services whose data you are willing to use for ad targeting are connected to Personal Intelligence.
- Disable unnecessary links: if you previously connected Gmail or Google Photos to Gemini or AI Mode but do not want ad personalization based on this data, disconnect the respective apps.
- Take residual data into account: according to Android Authority, disabling an app does not guarantee deletion of history already saved. Check and, if necessary, clear your Search Services History manually.
- For ChatGPT users: ads are shown only on the free tier and Go. If this is unacceptable, consider switching to a paid subscription.
Google’s policy update formalizes a compromise that until now has remained implicit: the convenience of AI personalization is paid for by expanding the advertising profile. The key action is to check your Personal Intelligence settings and disable cross‑service connections for Gmail, Photos and Drive if advertising use of this data is unacceptable, as well as to clear Search Services History to minimize the amount of information already collected.
